plotmcmc: plots output from 'atmcmc'

In atmcmc: Automatically Tuned Markov Chain Monte Carlo

Description

Produces traceplots and histograms of output from an ‘atmcmc’ run

Usage

plotmcmc(output, name = "MCMC", multimodal = F, plottype = "trace",
  format = "default", m = 10, phase.start = 1, phase.end = 5,
  nrow.trace = 3, ncol.trace = 1, nrow.hist = 3, ncol.hist = 3)

Arguments

output	output from an ‘atmcmc’ run
name	name of the project
multimodal	whether to assume target density function is strongly multimodal. Argument takes T or F
plottype	whether to produce traceplots or histograms, or both. Takes argument ‘trace’, ‘hist’, or ‘all’
format	file format. R graphics, pdf, or png. Takes argument ‘default’, ‘pdf’, or ‘png’
m	every ‘m’th iteration is plotted. Has to be a factor of ‘batchwidth’/2 from the ‘atmcmc’ run
phase.start	first phase to be plotted in traceplots. Takes argument 1, 2, 3, 4, or 5. 1 = 1st adaption phase, 2 = transient phase, 3 = 2nd adaption phase, 4 = 1st half of sampling phase, and 5 = 2nd half of sampling phase
phase.end	last phase to be plotted in traceplots. Takes argument 1, 2, 3, 4, or 5. 1 = 1st adaption phase, 2 = transient phase, 3 = 2nd adaption phase, 4 = 1st half of sampling phase, and 5 = 2nd half of sampling phase
nrow.trace	number of rows for traceplots per page
ncol.trace	number of columns for traceplots per page
nrow.hist	number of rows for histograms per page
ncol.hist	number of columns for histograms per page

Details

Histograms and traceplots are drawn for each coordinate separately. For the sampling phase, traceplots show only one replicative chain (chain that takes the last value of the 2nd adaption phase as the starting value of the sampling phase).

Value

Plots including

‘name’_histogram	histograms of the sample obtained from the ‘atmcmc’ run
‘name’_traceplot	traceplots from the ‘atmcmc’ run. Every ‘m’th iteration is plotted

Examples

## see examples in `atmcmc'

atmcmc documentation built on May 2, 2019, 11:05 a.m.